An Optimization Method for Updating Complex State Information in UAV-Assisted Internet of Things

Abstract

The present invention discloses a method for optimizing the update of complex state information in a drone-assisted Internet of Things. An optimization problem is established under the derived lower bounds of the average peak age of information, the average energy consumption of Internet of Things devices, and the average energy consumption of drones; according to the characteristics of the optimization variables and the problem structure, the original optimization problem is divided into five sub-problems, and an optimization method is proposed to alternately solve the five sub-problems. In the optimization method proposed in the present invention, the greedy algorithm solves the optimization problems of the access strategy and the state update strategy with low complexity; the successive convex approximation method can effectively solve the complex non-linear problems of the block length and the drone position optimization; the optimal value of the complex state information update request rate is given in a closed-form solution. The optimization method proposed in the present invention can obtain a sub-optimal solution to the original optimization problem.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication status information update, and particularly to an optimization method for complex status information update in a drone-assisted Internet of Things. Background Art

[0002] With the stable development of the Internet of Things (IoT), some emerging IoT applications (such as industrial IoT and health monitoring) need to continuously detect and update physical quantities (status information) in the real environment. During status update, the freshness of the status information should be ensured, and the Age of Information (AoI) proposed by S.K. Kaul et al. (S.K. Kaul, R.D. Yates, and M. Gruteser, “Real-time status: How often should one update?” in Proc. IEEE INFOCOM, 2012, pp. 2731–2735.) is an important performance metric reflecting the freshness of information. Existing literature mainly studies traditional status updates that only involve the communication process. However, the status information of some emergency applications (such as wildfire spread prediction and fire gas monitoring) needs to be obtained through complex calculations of raw data, and the status information needs to be transmitted to the control terminal for subsequent processing. Therefore, the present invention defines the status update process in which the status information needs to be obtained through complex calculations of raw data as complex status information update. However, energy-limited IoT devices may have difficulty bearing the energy consumption generated by executing computational tasks during complex status information update. Multi-access Edge Computing (MEC) allows IoT devices to offload computational tasks to nearby ground edge servers, thus effectively reducing the energy consumption of IoT devices. In recent years, drone-assisted emergency communication has become an effective communication method. Drones equipped with computing servers can help IoT devices complete complex status information update in emergency situations. Therefore, the present invention aims to study complex status information update in a drone-assisted Internet of Things.

[0003] Implementing complex status information update in a drone-assisted Internet of Things requires considering three performance metrics. First, the peak AoI can reflect the minimum freshness of the status information during complex status information update. Second, since the duration of some emergency applications (such as wildfire spread prediction) is not short, and even longer than the lifespan of IoT devices. Therefore, the energy consumption of IoT devices is an important performance metric. Third, since drones are also energy-limited devices, the energy consumption of drones must also be considered. In addition, due to the small amount of data collected by IoT devices, short-packet communication is used for communication between IoT devices and drones. In short-packet communication, the block length (i.e., a metric for the time-frequency resources used in communication, expressed as transmission time * channel bandwidth, with the unit of channel use) is limited. In the case of limited block length, the following two issues need to be considered when optimizing performance metrics.

[0004] First, with finite block lengths, the packet error rate (PER) is non-zero. Therefore, packet loss events may occur randomly during complex state information updates, resulting in random instantaneous values for the three performance metrics. To effectively optimize the three performance metrics, it is crucial to statistically derive their average values. Second, the three performance metrics are related to both the offloading process and the drone deployment process. The access policy, state update policy, block length, and complex state information update request rate are relevant to the offloading process and impact the performance metrics. During drone deployment, drone location affects the PER, which in turn affects the three performance metrics. Therefore, joint optimization of these variables should be considered. However, the PER is a complex, nonlinear function of block length and drone location. Furthermore, the access policy, state update policy, and block length are all integer variables. Therefore, the joint optimization of these multiple variables is a challenging mixed-integer nonlinear problem. Therefore, achieving effective joint optimization of these multiple variables with finite block lengths is a challenge. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for optimizing the update of complex state information in a drone-assisted Internet of Things for short packet communication scenarios, and to effectively optimize multiple variables under the described performance indicator framework to achieve efficient complex state information update in the drone-assisted Internet of Things.

[0006] The technical solution to achieve the purpose of the present invention is to provide a method for optimizing the update of complex state information in a drone-assisted Internet of Things, comprising the following steps:

[0007] Step 1: Establish the optimization problem of updating complex state information in drone-assisted IoT;

[0008] Step 2: Obtain feasible solutions for optimization variables through search methods n 0 ,L 0 ,λ 0 ,in, is a set of feasible access strategy variables, Update the set of policy variables for feasible states, n 0 is the feasible block length, L 0 is the set of feasible UAV positions, λ 0 is the feasible complex state information update request rate, and at the same time, let l=0,N max =20;

[0009] Step 3: In the feasible solution n l ,L l ,λ l Use the greedy algorithm to obtain a better access strategy set

[0010] Step 4: Obtain a better set of state update strategies using the greedy algorithm under the feasible solution n l , L l , λ l

[0011]

[0012] Step 5: Obtain a better block length n using the successive convex approximation method under the feasible solution n l , L l , λ l l+1 ;

[0013] Step 6: Obtain a better set of UAV positions L using the successive convex approximation method under the feasible solution n l+1 , L l , λ l l+1 ;

[0014] Step 7: Obtain the optimal complex state information update request rate λ according to n l+1 , L l+1 , λ l under the feasible solution, and let λ * , and let λ l+1 = λ * ;

[0015] Step 8: Let the value of l increase by 1, and repeat the above steps starting from Step 3 until l > N max . Where N max is the maximum number of iterations;

[0016] Step 9: Let n l , L l , λ l be the optimal solution for the complex state information update.

[0017] The complex state information update optimization method of the present invention constructs a joint optimization problem based on the derived analytical expressions of the lower bounds of the average peak AoI, average IoT device energy consumption, and average UAV energy consumption. This problem aims to minimize the weighted sum of the above three performance metrics by jointly optimizing the variables related to the offloading process and the UAV deployment process. Then, by analyzing the structure of this difficult joint optimization problem, the problem is transformed into multiple sub-problems, and each sub-problem is solved one by one to obtain a sub-optimal solution to the original joint optimization problem. ​​​​

[0018] Compared with the prior art, the significant advantages of the present invention are as follows: (1) simultaneously taking the peak AoI, the energy consumption of IoT devices, and the energy consumption of UAVs as performance indicators, enabling the entire network to achieve complex state information updates with high freshness under lower energy consumption of IoT devices and UAVs; (2) analyzing the performance indicators of the complex state information update process in UAV-assisted IoT, laying a theoretical foundation for network performance improvement; (3) in steps 2 and 3, the present invention proposes greedy algorithms for access strategy optimization and state update strategy optimization respectively to achieve a trade-off between performance and solution speed; (4) in steps 4 and 5, the present invention designs successive convex approximation methods for block length and UAV position respectively, effectively solving the optimization problems brought by the highly non-linear optimization objectives in block length optimization and UAV position optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 is a model diagram of the UAV-assisted IoT of the present invention.

[0020] Figure 2 is a schematic diagram of two state update strategies of the IoT devices of the present invention.

[0021] Figure 3 is a schematic diagram of the weighted sum of three performance indicators changing with the number of iterations of the present invention.

[0022] Figure 4 is a schematic diagram of the average peak age of information changing with the transmission power of IoT devices of the present invention.

[0023] Figure 5 is a schematic diagram of the average energy consumption of IoT devices changing with the transmission power of IoT devices of the present invention.

[0024] Figure 6 is a schematic diagram of the average energy consumption of UAVs changing with the transmission power of IoT devices of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0025] Figure 1 is a model diagram of the UAV-assisted IoT of the present invention. In UAV-assisted IoT, multiple IoT devices need to complete complex state information updates. The complex state information update process is as follows: IoT devices obtain raw data at a constant rate by sensing the environment and convert the raw data into state information through calculation, and this state information needs to be processed subsequently on the UAV. In the complex state information update process of IoT devices, multiple UAVs can act as computing servers to help IoT devices convert raw data into state information, thereby reducing the computing load and energy consumption pressure of IoT devices.

[0026] Figure 2It is a schematic diagram of two state update strategies for the Internet of Things devices of the present invention. To make full use of the resources of Internet of Things devices and drones, the Internet of Things devices have two complex state information update strategies, namely, the local computing and transmission strategy and the computing offloading strategy. In the local computing and transmission strategy, each Internet of Things device acquires raw data through sensors and executes its tasks locally, and then transmits the state information to the drone control terminal for subsequent processing. In the computing offloading strategy, each Internet of Things device offloads its raw data to a drone, and the drone acquires the state information by executing tasks for subsequent processing.

[0027] An embodiment of the present invention provides a method for optimizing the update of complex state information in a drone-assisted Internet of Things. The method includes the following steps:

[0028] Step 1: Establish an optimization problem for the update of complex state information in a drone-assisted Internet of Things. The optimization problem is as follows:

[0029]

[0030]

[0031]

[0032]

[0033]

[0034]

[0035]

[0036] n min ≤n≤n max , (1g)

[0037]

[0038]

[0039] where α k,m is the access strategy variable between the Internet of Things device k and the drone m. If the Internet of Things device k accesses the drone m, α k,m = 1. Otherwise, α k,m = 0. β k is the state update strategy variable of the Internet of Things device k. If β k = 1, then the Internet of Things device k selects the local computing and transmission strategy as the state update strategy. Otherwise, the Internet of Things device k selects the computing offloading strategy as the state update strategy. is the set of Internet of Things devices and K is the number of Internet of Things devices, Let \(\mathcal{U}\) be the set of UAVs and \(M\) be the number of UAVs, \(H\) min and \(H\) max are the minimum and maximum flight altitudes of the UAVs respectively, \(l\) x and \(l\) y are the three - dimensional coordinates of UAV \(x\) and UAV \(y\) respectively, \(d\) min is the minimum safety distance between two UAVs, \(W\) is the bandwidth allocated to each IoT device, \(f\) k is the computing frequency of IoT device \(k\), \(F\) is the computing frequency of any UAV, \(c\) k is the number of CPU revolutions required to convert the original data into state information during the update process of the complex state information of IoT device \(k\). Let \(\mathbb{Z}\) be the set of integers, \(n\) min and \(n\) max are the minimum and maximum block lengths in short - packet communication respectively. is the packet error probability when IoT device \(k\) selects the local computing and transmission strategy and accesses UAV \(m\). Among them, \(D_{si}\) k is the amount of state information data of IoT device \(k\). is the packet error probability when IoT device \(k\) selects the computing offloading strategy and accesses UAV \(m\). \(D_{ori}\) k is the amount of original data of IoT device \(k\). is the signal - to - interference - plus - noise ratio between IoT device \(k\) and UAV \(m\), \(p\) k is the transmission power of IoT device \(k\), \(\eta\) LoS and \(\eta\) NLoS are the additional path losses of the line - of - sight link and non - line - of - sight link respectively. \(f\) c is the carrier center frequency, \(c\) is the speed of light, \(N_0\) is the noise power spectral density. is the line - of - sight link probability between IoT device \(k\) and UAV \(m\). Among them, \(a\) and \(b\) are constants depending on the environment. In addition, \(l\) m and \(u\) k are the three - dimensional coordinates of UAV \(m\) and IoT device \(k\) respectively, \(l_z\) m is the \(Z\) - axis coordinate of UAV \(m\).

[0040] In Equation (1), is the weight factor of the peak information age, is the average peak information age of IoT device \(k\) accessing UAV \(m\). Among them,

[0041]

[0042]

[0043]

[0044] Among them,

[0045] In formula (1), is the weight factor of the energy consumption of the Internet of Things devices. The average energy consumption of the Internet of Things device k accessing the drone m is:

[0046]

[0047] Among them, κ is a constant related to the Internet of Things and drone computing architectures.

[0048] In formula (1), is the weight factor of the energy consumption of the drone, is the average value of the lower bound of the total energy consumption of all drones, and its expression is as follows:

[0049]

[0050] Among them, P H is the hovering power of the drone;

[0051] Step 2: Obtain a feasible solution of the optimization variable through a search method n 0 , L 0 , λ 0 . Among them, is the set of feasible access strategy variables, is the set of feasible state update strategy variables, n 0 is the feasible block length (i.e., a metric of the time-frequency resources used for communication, expressed as transmission time * channel bandwidth, with the unit of channel use), L 0 is the set of feasible drone positions, λ 0 is the feasible complex state information update request rate. At the same time, let l = 0, N max = 20;

[0052] Step 3: Obtain a better set of access strategies using the greedy algorithm under the feasible solution n l , L l , λ l Specifically, it includes the following steps: Specifically, it includes the following steps:

[0053] Step 3.1: Input the feasible solution n l , L l , λ l ;

[0054] Step 3.2: Calculate the objective function based on the feasible solution and Equation (1), and denote it as g global ;

[0055] Step 3.3: Let k = 1;

[0056] Step 3.4: Let m = 1;

[0057] Step 3.5: Determine whether the IoT device k accessing the drone m satisfies Equation (1a) to Equation (1i). If it satisfies, calculate the objective function value based on Equation (1) and this access strategy and other feasible solutions, and denote it as If it does not satisfy, do nothing;

[0058] Step 3.6: Let the value of m increase by 1 and start from Step 3.5 until m > M;

[0059] Step 3.7: After the IoT device k makes judgments on each drone, obtain as the set of all drones that satisfy Equation (1a) to Equation (1i). If is a non-empty set and Let and obtain a better access strategy for the IoT device k, that is where,

[0060] Step 3.8: Let the value of k increase by 1 and start from Step 3.4 until k > K;

[0061] Step 3.9: Output the set of better access strategies

[0062] Step 4: Use the greedy algorithm to obtain a set of better state update strategies under the feasible solution n l ,L l ,λ l ; Step 4 specifically includes Step 4.1 to Step 4.6;

[0063] Step 4.1: Input the feasible solution n l ,L l ,λ l ;

[0064] Step 4.2: Calculate the objective function g based on the feasible solution and Equation (1) global ;

[0065] Step 4.3: Let k = 1;

[0066] Step 4.4: If β k = 1, determine when βk Check whether equations (1a) to (1i) are satisfied when β = 0. If satisfied, calculate the objective function based on equation (1), β = 0, and other feasible solutions, and denote it as k Judge whether it holds. If it holds, let and obtain a better state update strategy for the Internet of Things device k, that is, β k = 0; if β k = 0, judge whether equations (1a) to (1i) are satisfied when β k = 1. If satisfied, calculate the objective function based on equation (1), β k = 1, and other feasible solutions, and denote it as Judge whether it holds. If it holds, let and obtain a better state update strategy for the Internet of Things device k, that is, β k = 1. ;

[0067] Step 4.5: Increase the value of k by 1 and start from Step 4.4 until k > K;

[0068] Step 4.6: Output the set of better state update strategies

[0069] Step 5: Obtain a better block length n using the successive convex approximation method under the feasible solution n l , L l , λ l l+1 ,

[0070] Step 5 specifically includes Steps 5.1 to 5.7;

[0071] Step 5.1: Input the feasible solution n l , L l , λ l and set the maximum number of iterations to 5;

[0072] Step 5.2: Let n local = n l , count = 1;

[0073] Step 5.3: Based on n local and the interior point method, solve the optimal solution n of the following convex approximation problem * ;

[0074]

[0075]

[0076] ​​

[0077] where ε max is the maximum tolerable error packet probability. The other parameter expressions for problem (5) are as follows:

[0078]

[0079]

[0080]

[0081]

[0082]

[0083]

[0084]

[0085]

[0086] where is when n = n local and other variables are input feasible solutions

[0087] where is when n = n local and other variables are input feasible solutions

[0088] where

[0089] where

[0090] Step 5.4: Let n local = n * , and increase the value of count by 1;

[0091] Step 5.5: Execute from Step 5.3 until count is greater than the maximum number of iterations;

[0092] Step 5.6: Search for the integer solution n * that is closest to n IS and satisfies equations (1b) to (1i), and let n l+1 = n IS;

[0093] Step 5.7: Output a better block length n l+1 .

[0094] Step 6: Obtain a better set of UAV positions L using the sequential convex approximation method under the feasible solutions n l+1 , L l , λ l , specifically including the steps: l+1 .

[0095] Step 6.1: Input the feasible solutions n l+1 , L l , λ l and set the maximum number of iterations to 5;

[0096] Step 6.2: Let L local = L l , count = 1;

[0097] Step 6.3: Based on L local and the interior point method, solve for the optimal solution L of the following convex approximation problem * ;

[0098]

[0099]

[0100]

[0101]

[0102]

[0103]

[0104]

[0105]

[0106] where d k,m , ρ k,m , θ k,m , are the introduced auxiliary optimization variables, is the three-dimensional coordinate of UAV m when L = L local , and are the X-axis coordinate, Y-axis coordinate, and Z-axis coordinate of UAV m when L = L local , respectively. ux k and uy k are the X-axis coordinate and Y-axis coordinate of the Internet of Things device k, respectively. The other parameter expressions of question (6) are as follows:

[0107]

[0108]

[0109]

[0110]

[0111]

[0112]

[0113] Among them, is when L = L local and other variables are input feasible solutions At the same time, Among them, and

[0114] Among them, is when L = L local and other variables are input feasible solutions At the same time, Among them,

[0115]

[0116] Among them,

[0117]

[0118] Among them,

[0119]

[0120] Among them, is the i-th sample point among 10 equally spaced sampled points within the value range of ρ determined by equations (6d) and (6g). k,m Take 10 4 equally spaced sampled points within the value range of ρ determined by equations (6d) and (6g).

[0121] The maximum value can first be obtained by searching for those that satisfy of and then calculating Meanwhile, wherein,

[0122] wherein, The maximum value can first be obtained by searching for those that satisfy of and then calculating Meanwhile, wherein,

[0123]

[0124] Step 6.4: Let L local = L * and increment count by 1;

[0125] Step 6.5: Restart from Step 6.3 until count is greater than the maximum number of iterations;

[0126] Step 6.6: Output the better UAV position L l+1 .

[0127] Step 7: Optimization of the complex state information update request rate: Obtain the optimal complex state information update request rate λ n l+1 , L l+1 , λ l under and let λ * = λ l+1 = λ * ;

[0128] Step 8: Increment the value of l by 1 and repeat the above steps starting from Step 3 until l > N max ;

[0129] Step 9: Let n l , L l , λ l be the optimized solution for the complex state information update.

[0130] Example:

[0131] Simulation parameter settings: The side length of the square area where the IoT devices are located is 200 meters, the number of IoT devices K = 20, the transmission power of the IoT devices is p k = 10 dBm, and the computing frequency of the IoT devices is f k= 0.7 GHz, the bandwidth allocated to each IoT device is 2 MHz, the number of drones M = 4, the computing frequency of the drones is F = 2.6 GHz, the minimum flight height H of the drones min = 100 meters, the maximum flight height H of the drones max = 150 meters, the minimum safe distance between drones is d min = 10 meters, the minimum block length is n min = 100 channel use, the maximum block length is n max = 1000 channel use, the constant κ related to the computing architecture is 10 -28 , the data volume of the original data is Dori k = 32 bits, the data volume of the status information is Dsi k = Dori k - 2 bytes, the maximum tolerable packet error probability is ε max = 10 -5 , the environmental constants a = 11.95, b = 0.14, the carrier center frequency f c = 2.4 GHz, the noise power spectral density is N0 = - 174 dBm / Hz. The weight of the drone is about 2 KG, and its hovering power is about P H = 168.5 W (Y. Zeng, J. Xu, and R. Zhang, “Energy minimization for wireless communication with rotary - wing UAV,” IEEE Transactions on Wireless Communications, vol. 18, no. 4, pp. 2329–2345, Apr. 2019). In addition, the number of Monte Carlo simulation times to obtain the true value is 10 4 .

[0132] Figure 3 is a schematic diagram of the weighted sum of three performance indicators of the present invention changing with the number of iterations. It can be seen from the figure that under different original data volumes, the optimization method proposed in the present invention can converge within a limited number of iterations, reflecting the good convergence performance of the proposed optimization method. At the same time, the smaller the original data volume, the better the performance of the proposed optimization method. This is because the smaller the original data volume, the smaller the computational delay and energy consumption, thus reducing the weighted sum of the three performance indicators. At the same time, when the original data volume decreases, the convergence speed of the proposed optimization method is faster. This shows that the original data volume will affect the convergence speed of the proposed optimization method and the proposed method is more suitable for optimizing emergency applications with smaller original data volumes.

[0133] Figure 4 , Figure 5 andFigure 6 Schematic diagrams of the peak information age, the energy consumption of IoT devices, and the energy consumption of UAVs of the present invention varying with the transmission power of IoT devices, respectively. From Figure 4 it can be found that, except for "grid search", the optimization method proposed in the present invention is superior to other methods and approaches the performance of "grid search" at a lower transmission power, which indicates that the proposed optimization method can effectively reduce the average peak information age. In addition, since the optimization method only obtains a suboptimal solution and there are often more than one suboptimal solution, fluctuations can be observed in the curve corresponding to the proposed optimization method. In Figure 5 and Figure 6 , the proposed optimization method is superior to the "fixed block length" method and the "initial solution", and has the same performance as the "fixed complex state information update request rate", which indicates that the proposed optimization method can effectively reduce the energy consumption of IoT devices and UAVs, and the complex state information update request rate has no impact on the energy consumption of IoT devices and UAVs. At the same time, combining Figure 4 , Figure 5 and Figure 6 it can be seen that, since the monotonicity of the three performance metrics with respect to the transmission power of IoT devices is inconsistent, it is necessary to select an appropriate transmission power of IoT devices to improve the network performance.

[0134] The description of the above embodiments is relatively specific and detailed, but it only represents a feasible implementation manner of the present invention and does not limit the scope of the present invention patent. It should be noted that scientific researchers and engineers in the field can add several deformations or improvements on the basis of this embodiment within the framework of the present invention, but these are all within the protection scope of the present invention patent, and the protection scope of the present invention patent shall be subject to the appended claims.

Claims

1. An optimization method for updating complex state information in a drone-assisted Internet of Things, characterized in that, Including the following steps: Step 1: Establish the problem of optimizing the update of complex state information in a UAV-assisted Internet of Things; Step 2: Optimize variable initialization: Obtain the feasible solutions of the optimization variables through a search method Among them, is the set of feasible access strategy variables, is the set of feasible state update strategy variables, n 0 is the feasible block length, L 0 is the set of feasible UAV positions, λ 0 is the feasible complex state information update request rate; meanwhile, let l = 0, N max = 20; Step 3: Access policy variable optimization: Obtain a more optimal access policy set using the greedy algorithm under the feasible solution ​ Step 4: State update strategy optimization: Under the feasible solution obtain a more optimal set of state update strategies using the greedy algorithm Step 5: Block length optimization: Obtain a better block length n by using the sequential convex approximation method under the feasible solution ; l+1 ; Step 6: UAV position optimization: Obtain a more optimal set of UAV positions L by using the sequential convex approximation method under the feasible solution ; l+1 ​ Step 7: Optimize the request rate for complex state information updates: Based on Obtain the optimal complex state information update request rate λ * , and let λ l+1 =λ * ;W is the bandwidth allocated to each IoT device, α k,m is the access strategy variable between IoT device k and drone m, F is the calculation frequency of any drone, c k is the number of CPU revolutions required to convert raw data into status information during the complex status information update process of IoT device k, Step 8: Increase the value of l by 1 and repeat the above steps starting from Step 3 until l > N max ; where N max is the maximum number of iterations; Step 9: Let be the optimized solution for complex state information update.

2. A method for updating and optimizing complex state information in a drone-assisted Internet of Things according to claim 1, characterized in that, To make full use of the resources of Internet of Things devices and UAVs, the Internet of Things devices have two complex state information update strategies, namely the local computing and transmission strategy and the computing offloading strategy; in the local computing and transmission strategy, each Internet of Things device acquires raw data through sensors and executes its tasks locally, and then transmits the state information to the UAV control terminal for subsequent processing; in the computing offloading strategy, each Internet of Things device offloads its raw data to a UAV, and the UAV obtains the state information by executing tasks for subsequent processing.

3. A method for updating and optimizing complex state information in an unmanned aerial vehicle-assisted Internet of Things according to claim 2, characterized in that, The optimization problem described in Step 1 is as follows: Among them, α k,m is the access strategy variable between IoT device k and drone m. If IoT device k accesses drone m, α k,m =1; otherwise, α k,m =0;β k Update the policy variable for the state of IoT device k. If β k =1, the IoT device k selects the local computation transmission strategy as the state update strategy; otherwise, the IoT device k selects the computation offloading strategy as the state update strategy; is the set of IoT devices and K is the number of IoT devices, is the set of drones and M is the number of drones, H min and H max are the minimum and maximum flight altitudes of the UAV, respectively. x and l y are the three-dimensional coordinates of drone x and drone y, d min is the minimum safe distance between two drones, W is the bandwidth allocated to each IoT device, and f k is the computing frequency of IoT device k, F is the computing frequency of any drone, c k is the number of CPU revolutions required to convert raw data into status information during the complex status information update process of IoT device k, is a set of integers, n min and n max They are the minimum block length and maximum block length in short packet communication respectively; The packet error probability when the local computing transmission strategy is selected for IoT device k and connected to drone m; where, Dsi k is the amount of status information data of IoT device k, The packet error probability when selecting the computation offloading strategy for IoT device k and connecting to drone m, Dori k is the amount of raw data of IoT device k; is the signal-to-interference-noise ratio between IoT device k and drone m, p k is the transmission power of IoT device k, η LoS and η NLoS The additional path losses for line-of-sight and non-line-of-sight links, f c is the carrier center frequency, c is the speed of light, N0 is the noise power spectrum density, is the probability of a line-of-sight link between IoT device k and UAV m; where a and b are constants that depend on the environment; l m and u k They are the three-dimensional coordinates of the drone m and the Internet of Things device k, respectively, lz m is the Z-axis coordinate of the drone m; In formula (1), is the weight factor of the peak information age, is the average peak information age of the Internet of Things device k accessing the drone m; where Among them, In formula (1), is the weight factor of the energy consumption of the Internet of Things device. The average energy consumption of the Internet of Things device k connected to the drone m is: where κ is a constant related to the computing architectures of the Internet of Things and UAVs; In formula (1), is the weight factor of the energy consumption of the UAV, is the average value of the lower bound of the total energy consumption of all UAVs, and its expression is as follows: Among them, P H is the hovering power of the drone.

4. A method for updating and optimizing complex state information in an unmanned aerial vehicle-assisted Internet of Things according to claim 3, characterized in that The specific process of the greedy algorithm described in Step 3 is as follows: Step 3.1: Input the feasible solution Step 3.2: Calculate the objective function based on the feasible solution and Equation (1), and denote it as g global ; Step 3.3: Let k = 1; Step 3.4: Let m = 1; Step 3.5: Determine whether the IoT device k satisfies equations (1a) to (1i) when accessing the drone m. If it is satisfied, calculate the objective function value based on equation (1), this access policy, and other feasible solutions, and denote it as If it is not satisfied, do nothing; Step 3.6: Increase the value of m by 1 and restart from Step 3.5 until m > M; Step 3.7: After IoT device k makes judgments on each drone, it obtains the set of all drones that satisfy equations (1a) to (1i); if is a non-empty set and let and obtain a better access strategy for IoT device k, that is where Step 3.8: Increase the value of k by 1 and restart from Step 3.4 until k > K; Step 3.9: Output a more optimal set of access policies 5. A method for updating and optimizing complex state information in a drone-assisted Internet of Things according to claim 4, characterized in that, The specific process of the greedy algorithm described in Step 4 is as follows: Step 4.1: Input the feasible solution Step 4.2: Calculate the objective function g based on the feasible solution and Equation (1) global ; Step 4.3: Let k = 1; Step 4.4: If β k = 1, determine whether equations (1a) to (1i) are satisfied when β k = 0; if satisfied, calculate the objective function based on equation (1), β k = 0, and other feasible solutions, and denote it as Determine whether holds; if it holds, let and obtain a better state update strategy for the Internet of Things device k, that is, β k = 0; if β k = 0, determine whether equations (1a) to (1i) are satisfied when β k = 1; if satisfied, calculate the objective function based on equation (1), β k = 1, and other feasible solutions, and denote it as Determine whether holds; if it holds, let and obtain a better state update strategy for the Internet of Things device k, that is, β k = 1; Step 4.5: Increase the value of k by 1 and start from Step (4.4) until k > K; Step 4.6: Output a set of more optimal state update strategies 6. A method for updating and optimizing complex state information in a drone-assisted Internet of Things according to claim 5, characterized in that The specific process of the successive convex approximation method described in Step 5 is as follows: Step 5.1: Input a feasible solution And set the maximum number of iterations to 5; Step 5.2: Let n local = n l , count = 1; Step 5.3: Based on n local and the interior point method, solve for the optimal solution n of the following convex approximation problem * ; where ε max is the maximum tolerable error packet probability; the other parameter expressions of formula (5) are as follows: Among them, when n = n local and other variables are input feasible solutions Among them, when n = n local and other variables are input feasible solutions Among them, Among them, Step 5.4: Let n local = n * , and increment the value of count by 1; Step 5.5: Execute from Step 5.3 until count is greater than the maximum number of iterations; Step 5.6: Search for the integer solution n that is closest to n * and satisfies the equations (1b) to (1i) IS and let n l+1 = n IS ; Step 5.7: Output a better block length n l+1 .

7. A method for updating and optimizing complex state information in a drone-assisted Internet of Things according to claim 6, characterized in that, The specific process of the successive convex approximation method described in Step 6 is as follows: Step 6.1: Input a feasible solution And set the maximum number of iterations to 5; Step 6.2: Let L local = L l , count = 1; Step 6.3: Based on L local and the interior point method, solve for the optimal solution L of the following convex approximation problem * ; Among them, is the introduced auxiliary optimization variable, is the three-dimensional coordinates of the unmanned aerial vehicle m when L = L local ; and are respectively the X-axis coordinate, Y-axis coordinate and Z-axis coordinate of the unmanned aerial vehicle m when L = L local ; ux k and uy k are respectively the X-axis coordinate and Y-axis coordinate of the Internet of Things device k; the other parameter expressions of formula (6) are as follows: Among them, is when L = L local and other variables are input feasible solutions At the same time, Among them, and Among them, when L = L local and other variables are input feasible solutions At the same time, Among them, Among them, Among them, Among them, is the i-th sample point among 10 equally spaced sampled points within the value range of ρ determined by formulas (6d) and (6g); k,m Take 10 4 equally spaced sampled points within the value range of ρ; This maximum value can first be obtained by searching for those that satisfy and then calculating At the same time, Among them, Among them, Among them, This maximum value can first be obtained by searching for those that satisfy of and then calculating Meanwhile, Among them, Step 6.4: Let L local = L * and increment count by 1; Step 6.5: Restart from Step 6.3 until count is greater than the maximum number of iterations; Step 6.6: Output the better drone position L l+1 .

Images